Misconduct session
Misconduct is what most people call “fraud”. This session had three speakers.
The first speaker was the Editor-in-Chief of the journal Anaesthesia, Steve Yentis. Yentis told about the case of Joachim Boldt, an anesthesiologist who has had over 80 papers retracted. He also told about the case of case of Yoshitaka Fujii, an anesthesiologist who seems to have published 193 bogus papers. A third case was also cited, though I did not get the details. Yentis has been leading the charge to get more integrity in anesthesiology.
The second speaker was Peter Aldhous, from New Scientist magazine. Aldhous has worked to expose fraud with stories in New Scientist. His presentation seemed sound, but there was no substantial news. (The slides for his presentation are on his web site.) One point he made was that institutions are unlikely to fairly investigate allegations of fraud made against their own researchers: this obvious point seemed to be new to some people.
The third speaker was Ginny Barbour, who is the Chair of the Committee On Publication Ethics (COPE). Barbour said that COPE was working to get institutions to investigate allegations of fraud made against their own researchers. She also claimed that only a few percent of research publications are fraudulent.
During the question period, someone stated that science journalists should be cheerleading science, and that fraud is very rare, and anyway science is self-correcting. More generally, many people there genuinely believed that almost all scientists are virtually always honest. Those people work with science all day, and yet they seem to have no clue about how science really operates. Overall, I found the session stunningly disheartening: there is an enormous way to go, to get many journalists to appreciate what reality is.
I pointed out that all the examples of fraud given by the speakers were in medical science. I noted that in the UK, during the past half century, there does not seem to have been a single case where a non-medical researcher has been officially found to have committed fraud. That is clearly unreasonable. Consider much smaller groups of respected people: members of parliament, Catholic priests, police detectives—in each instance, we know that during half a century, at least a few of them will have committed serious crimes.
I also described how I once reported a fraud at the University of Reading. The university refused to investigate: I was told that the university had no procedures for investigating such allegations, because their professors always act with integrity.
The conclusion is that there is no accountability. I said that there were some fields of science where half the research publications were bogus. That was in conflict with the claim of Barbour, and did not go over well.
Some journalists seemed to think that verifying research fraud requires specialist scientific expertise. I gave two counterexamples from my own work. One counterexample is in archaeoastronomy of China, where a fraud consisted of claiming that a figure with four dots in it actually had five dots in it: in other words, understanding the fraud only requires being able to count to five. I published a paper on this, which overturned the previous 20 years of research in the field. The other counterexample was the analysis by Phil Jones on Chinese weather stations; I mentioned this only briefly, as I did not want to get into the emotive politics of global warming.
Afterwards, I briefly talked with Aldhous. Aldhous explained how a journalist might have to put in as much time to get a story about fraud as to get, say, 75 stories about usual science. From a business perspective, that is obviously a serious problem.
I also briefly talked with Barbour. I repeated the point that Aldhous had made about institutions investigating their own researchers. Barbour replied that she was aware of the problem, but having institutions investigate their own was the best that could currently be done. At the time, I could not think of anything polite to say in response to such nonsense. It is obvious that having institutions investigate their own is worse than doing nothing, because it tends to give the illusion of there being a real investigation. (We have seen such illusory investigations with Climategate, for example.)
Barbour then talked about the COPE Code of Conduct and Best Practice Guidelines for Journal Editors. She suggested that it was via the Code that research would gain more integrity. The Code contains statements saying that editors should “strive to constantly improve their journal”, should be “supporting initiatives to educate researchers about publication ethics”, and should “publish guidance to authors on everything that is expected of them”. Those are plainly platitudes. The Code makes no mention of research data having to be disclosed, of computer programs having to be available, etc.—that is, it lacks most of the specifics that would be needed for it to accomplish anything non-illusory.
Plenary session
This session had four speakers.
The first speaker was the journalist William Cullerne Bown. Bown said that science journalism is failing: his main evidence is the small number of readers. He explained how, after the development of the atomic bomb, the public became fascinated by science and the promise to change civilization. That promise has not been fulfilled, at least not nearly as much as was claimed half a century ago. In consequence, public interest in science has lessened, and shifted more to technology.
Bown noted that the big stories about scientific failures have been missed. He also commented on large institutions involved in science—citing the Royal Society, Elsevier, and the Wellcome Trust—saying that such institutions “have their own interests”, which are not necessarily those of science. He observed that there is lots of corruption among politicians and business people; so how could it be that there is seemingly almost none among scientists? He criticized the many science journalists who seem to treat scientists as being almost like demigods.
The second speaker was the economist and BBC journalist Evan Davis. Davis cautioned journalists about focusing on negative aspects of science, such as fraud, saying that negative publicity might bias scientists against doing anything, and we want to encourage scientists to continue their work. He later said that “exposure [of fraud] is not a very good goal” for science journalism.
I think that there are many people like Davis: people who have effectively taken science to be their religion and scientists to be their priests or even gods. Those people are deeply fearful of having their religious beliefs defenestrated. The result is the anti-journalism on display here.
The third speaker was Jay Rosen, Associate Professor of Journalism at New York University. Rosen was also the person who delivered the keynote address for the conference. He stated that it was important for journalists to “confront climate change denialism”. He came across as being highly certain of himself, while having little understanding of the issues, i.e. a typical third-rate academic.
The fourth speaker was the President of the Association of British Science Writers, Connie St Louis; St Louis is also the Director of Science Journalism at City University London. St Louis argued that journalists are too close to scientists. She said that fraud is “very very underreported” and that in consequence “we have failed as journalists”. She closed with the exhortation “let’s have some real journalism”.
The statements by St Louis prompted discussion. Bown said that he agreed with the statements. Many people, however, seemed to disagree. Someone in the audience said that fraud is confined to a few isolated individuals.
I repeated my main point from the earlier session: in the UK, during the past half century, there does not seem to have been a single case where a non-medical researcher has been officially found to have committed fraud. This demonstrates that there is no accountability.
I also gave an example from my own work, in radiocarbon dating. In radiocarbon dating, a chemical measurement is made on the remains of an organism, and then a statistical procedure is used to calculate how many years ago the organism died. I had found an error in the statistical procedure: thus, most radiocarbon dates are inaccurate. I submitted a paper on this to a journal. The paper had five peer reviews, all recommending rejection; in each case, I wrote a rebuttal—in some cases pointing out clear dishonesty by the reviewer. The editors actually asked a total of 25 scientists to peer review the paper; the remaining 20 declined to do a review, but were often were critical of the paper’s claim. In other words, out of 25 scientists, not one could be found to recommend accepting the paper. Eventually, the journal sent the paper to a statistician, whom I was told was eminent; the statistician said that the paper was obviously correct. The paper was then published. Thus, this is an example of an entire field covering up a substantial error—presumably because they want to avoid the embarrassment of admitting making such a mistake.
___________________For me, the take-home message from the conference is that a large majority of science journalists are extremely naive about scientists. The naivety is so extreme that I suspect it must be partially willful.
For global-warming skeptics, something else should perhaps be mentioned. Many global-warming skeptics seem to think that there is something special about the prevalence of bogus research in global warming. There is not. Anyone who has looked at other fields of science knows that there are fields that are worse than global warming. This tells us something important: the underlying cause of the problem is not specific to global warming.
I mention this especially because some skeptics seem to believe that what is needed is reform of the IPCC. Yes, the IPCC could benefit from reform. But that would not solve the problem.
We have known for millennia that prerequisites for integrity in human affairs include things like transparency and accountability. Those things should be in all scientific research.